The experiments begun at Newfoundland were transferred to Nova Scotia, and at Glace Bay in 1902 was established a station from which messages were transmitted and experimental work carried on until its work was temporarily interrupted by fire in 1909.  Here four wooden lattice towers, each 210 feet in height, were built at the corner of a square 200 feet on a side, and a conical arrangement of 400 copper wires supported on stays between the tops of the towers and connected in the middle at the generating station was built.  Additional machinery was installed and at the same time a station at Cape Cod for commercial work was built.  In December, 1902, regular communication was established between Glace Bay and Poldhu, but it was only satisfactory from Canada to England as the apparatus at the Poldhu station was less powerful and efficient than that installed in Canada.  The transmission of a message from President Roosevelt to King Edward marked the practical beginning of trans-Atlantic wireless telegraphy.  By this time a new device for the detection of messages was employed, as the coherer we have described even in its improved forms was found to possess its limitations of sensitiveness and did not respond satisfactorily to long distance signals.  A magnetic detector was devised by Marconi while other inventors had contrived electrolytic, mercurial, thermal, and other forms of detector, used for the most part with a telephone receiver in order to detect minute variations in the current caused by the reception of the electro-magnetic waves.  With one of Marconi’s magnetic detectors signals from Cape Cod were read at Poldhu.

In 1903 wireless telegraphy had reached such a development that the transmission of news messages was attempted in March and April of that year.  But the service was suspended, owing to defects which manifested themselves in the apparatus, and in the meantime a new station in Ireland was erected.  But there was no cessation of the practical experiments carried on, and in 1903 the Cunard steamship Lucania received, during her entire voyage across from New York to Liverpool, news transmitted direct from shore to shore.  In the meantime intercommunication between ships had been developed and the use of wireless in naval operations was recognized as a necessity.

Various improvements from time to time were made in the aerial wires, and in 1905 a number of horizontal wires were connected to an aerial of the inverted cone type previously used.  The directional aerial with the horizontal wires was tried at Glace Bay, and adopted for all the long distance stations, affording considerable strengthening of the received signals at Poldhu stations.  Likewise improvements in the apparatus were effected at both trans-Atlantic stations, consisting of the adoption of air condensers composed of insulated metallic plate suspended in the air, which were found much better than the condensers where glass was previously used to separate the plates.  For producing the energy employed for transmitting the signals a high tension continuous current dynamo is used.  An oscillatory current of high potential is produced in a circuit which consists of rapidly rotating disks in connection with the dynamo and suitable condensers.